clip-vit-base-patch32

This model is a fine-tuned version of openai/clip-vit-base-patch32 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2471
• Accuracy: 0.9078
• Precision: 0.9039
• Recall: 0.8956
• F1: 0.8997
• Tp: 1467
• Tn: 1754
• Fp: 156
• Fn: 171

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0003
• train_batch_size: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 55
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Tp	Tn	Fp	Fn
0.5345	0.0991	11	0.6493	0.6928	0.6057	0.9585	0.7423	1570	888	1022	68
0.8227	0.1982	22	1.0054	0.6463	0.5680	0.9768	0.7183	1600	693	1217	38
0.7845	0.2973	33	0.3412	0.8512	0.9752	0.6954	0.8118	1139	1881	29	499
0.6277	0.3964	44	0.6888	0.6787	0.5932	0.9676	0.7355	1585	823	1087	53
0.7787	0.4955	55	1.4060	0.5984	0.9908	0.1313	0.2318	215	1908	2	1423
1.1497	0.5946	66	2.5901	0.5383	0.0	0.0	0.0	0	1910	0	1638
1.4769	0.6937	77	0.5430	0.7610	0.7297	0.7662	0.7475	1255	1445	465	383
0.9893	0.7928	88	0.6155	0.6105	0.9961	0.1569	0.2711	257	1909	1	1381
0.8512	0.8919	99	0.4783	0.7466	0.6588	0.9359	0.7733	1533	1116	794	105
0.5253	0.9910	110	0.5975	0.6178	0.9519	0.1813	0.3046	297	1895	15	1341
0.8223	1.0901	121	0.9666	0.6229	0.5516	0.9792	0.7057	1604	606	1304	34
0.6539	1.1892	132	0.5507	0.6919	0.6048	0.9603	0.7422	1573	882	1028	65
0.5392	1.2883	143	0.4230	0.7458	0.6567	0.9414	0.7737	1542	1104	806	96
0.4739	1.3874	154	0.3672	0.8565	0.8390	0.8529	0.8459	1397	1642	268	241
0.4530	1.4865	165	0.4269	0.7951	0.7166	0.9200	0.8057	1507	1314	596	131
0.4480	1.5856	176	0.3274	0.8923	0.9015	0.8608	0.8807	1410	1756	154	228
0.4566	1.6847	187	0.3487	0.8822	0.8827	0.8590	0.8707	1407	1723	187	231
0.4529	1.7838	198	0.3308	0.8577	0.7968	0.9286	0.8576	1521	1522	388	117
0.4565	1.8829	209	0.3224	0.8571	0.8985	0.7784	0.8342	1275	1766	144	363
0.4378	1.9820	220	0.3924	0.8151	0.7395	0.9255	0.8221	1516	1376	534	122
0.4249	2.0811	231	0.2724	0.9025	0.9449	0.8376	0.8880	1372	1830	80	266
0.4260	2.1802	242	0.3528	0.8393	0.8048	0.8608	0.8319	1410	1568	342	228
0.4263	2.2793	253	0.3342	0.8464	0.8041	0.8822	0.8413	1445	1558	352	193
0.4263	2.3784	264	0.3556	0.8388	0.7826	0.9011	0.8377	1476	1500	410	162
0.4040	2.4775	275	0.2807	0.9045	0.9220	0.8663	0.8933	1419	1790	120	219
0.4115	2.5766	286	0.3133	0.8546	0.8934	0.7778	0.8316	1274	1758	152	364
0.4242	2.6757	297	0.3528	0.8396	0.7958	0.8779	0.8348	1438	1541	369	200
0.4228	2.7748	308	0.3443	0.8546	0.8066	0.9011	0.8512	1476	1556	354	162
0.3898	2.8739	319	0.3491	0.8292	0.7557	0.9310	0.8342	1525	1417	493	113
0.4204	2.9730	330	0.3132	0.8890	0.9424	0.8089	0.8706	1325	1829	81	313
0.3840	3.0721	341	0.3697	0.8255	0.7606	0.9078	0.8277	1487	1442	468	151
0.3856	3.1712	352	0.3160	0.8653	0.8226	0.9029	0.8609	1479	1591	319	159
0.3864	3.2703	363	0.3139	0.8842	0.8602	0.8944	0.8770	1465	1672	238	173
0.3557	3.3694	374	0.3745	0.8357	0.7712	0.9158	0.8373	1500	1465	445	138
0.3643	3.4685	385	0.2845	0.9016	0.8966	0.8895	0.8930	1457	1742	168	181
0.3485	3.5676	396	0.2970	0.8794	0.8513	0.8950	0.8726	1466	1654	256	172
0.3550	3.6667	407	0.3434	0.8613	0.8435	0.8590	0.8512	1407	1649	261	231
0.3964	3.7658	418	0.2631	0.9166	0.9692	0.8462	0.9035	1386	1866	44	252
0.3196	3.8649	429	0.2642	0.9078	0.9456	0.8492	0.8948	1391	1830	80	247
0.3296	3.9640	440	0.2820	0.8971	0.9399	0.8303	0.8817	1360	1823	87	278
0.3239	4.0631	451	0.3410	0.8667	0.8161	0.9182	0.8641	1504	1571	339	134
0.3365	4.1622	462	0.2807	0.9005	0.9028	0.8791	0.8908	1440	1755	155	198
0.3164	4.2613	473	0.2186	0.9315	0.9401	0.9096	0.9246	1490	1815	95	148
0.2912	4.3604	484	0.2564	0.9076	0.9384	0.8559	0.8953	1402	1818	92	236
0.3235	4.4595	495	0.2432	0.9101	0.9681	0.8327	0.8953	1364	1865	45	274
0.3203	4.5586	506	0.2816	0.9005	0.8800	0.9084	0.8940	1488	1707	203	150
0.3154	4.6577	517	0.2376	0.9208	0.9347	0.8907	0.9122	1459	1808	102	179
0.2780	4.7568	528	0.2341	0.9132	0.9677	0.8400	0.8993	1376	1864	46	262
0.2997	4.8559	539	0.2340	0.9107	0.9126	0.8919	0.9021	1461	1770	140	177
0.2629	4.9550	550	0.2471	0.9078	0.9039	0.8956	0.8997	1467	1754	156	171

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2